ZZ revised this gist . Go to revision
1 file changed, 4 insertions, 2 deletions
Gemini-ZFS-Fix.md
| @@ -1,7 +1,9 @@ | |||
| 1 | - | - Root Cause 1 (Kernel Panic): An interface bug between Linux Kernel 6.18 code allocation profiling hooks (CONFIG_MEM_ALLOC_PROFILING) and the out-of-tree OpenZFS driver module. Background ZFS I/O workers drop pointers during slab memory cleanups, triggering __alloc_tagging_slab_free_hook page faults. | |
| 1 | + | When you have 128 GB of RAM, ZFS defaults to a 64 GB ARC max ceiling. If your Windows VM pins 96 GB via hugepages, your system is mathematically trying to commit 160 GB of RAM on a 128 GB physical machine. | |
| 2 | + | Since hugepages cannot be unpinned, the host kernel starves, leading directly to the brutal cascading page faults and segmentation faults you experienced. | |
| 2 | 3 | ||
| 4 | + | Technical Case Notes: | |
| 5 | + | - Root Cause 1 (Kernel Panic): An interface bug between Linux Kernel 6.18 code allocation profiling hooks (CONFIG_MEM_ALLOC_PROFILING) and the out-of-tree OpenZFS driver module. Background ZFS I/O workers drop pointers during slab memory cleanups, triggering __alloc_tagging_slab_free_hook page faults. | |
| 3 | 6 | - Root Cause 2 (Segfaults / Exit Code 139): Core memory allocation collision. Pinned Windows VM static hugepages (96 GB) combined with default OpenZFS ARC ceilings (50% of host RAM = 64 GB) exceeded physical RAM capacity (128 GB). This starved the host kernel during intensive parallel nixos-rebuild compiler routines, triggering segmentation faults across Python, Sphinx, and VS Code extraction tools. | |
| 4 | - | ||
| 5 | 7 | - Neutralize Kernel Allocation Profiling HooksForce the Linux kernel to bypass the unpatched memory tracking hooks that conflict with out-of-tree file system drivers.Action: Upgrade to a fixed minor kernel release (e.g., 6.18.48+) and append alloc_tag_boot=off directly to the boot loader command line parameters. | |
| 6 | 8 | ||
| 7 | 9 | ``` | |
ZZ revised this gist . Go to revision
1 file changed, 30 insertions
Gemini-ZFS-Fix.md(file created)
| @@ -0,0 +1,30 @@ | |||
| 1 | + | - Root Cause 1 (Kernel Panic): An interface bug between Linux Kernel 6.18 code allocation profiling hooks (CONFIG_MEM_ALLOC_PROFILING) and the out-of-tree OpenZFS driver module. Background ZFS I/O workers drop pointers during slab memory cleanups, triggering __alloc_tagging_slab_free_hook page faults. | |
| 2 | + | ||
| 3 | + | - Root Cause 2 (Segfaults / Exit Code 139): Core memory allocation collision. Pinned Windows VM static hugepages (96 GB) combined with default OpenZFS ARC ceilings (50% of host RAM = 64 GB) exceeded physical RAM capacity (128 GB). This starved the host kernel during intensive parallel nixos-rebuild compiler routines, triggering segmentation faults across Python, Sphinx, and VS Code extraction tools. | |
| 4 | + | ||
| 5 | + | - Neutralize Kernel Allocation Profiling HooksForce the Linux kernel to bypass the unpatched memory tracking hooks that conflict with out-of-tree file system drivers.Action: Upgrade to a fixed minor kernel release (e.g., 6.18.48+) and append alloc_tag_boot=off directly to the boot loader command line parameters. | |
| 6 | + | ||
| 7 | + | ``` | |
| 8 | + | # Disable hugepages before recompiling kernel | |
| 9 | + | sudo sysctl -w vm.nr_hugepages=0 | |
| 10 | + | sudo sync && echo 3 | sudo tee /proc/sys/vm/drop_caches | |
| 11 | + | sudo echo never | sudo tee /sys/kernel/mm/transparent_hugepage/enabled | |
| 12 | + | ``` | |
| 13 | + | ||
| 14 | + | ZFS Notes | |
| 15 | + | - ZFS ARC Mem Use: `zarcstat # size` | |
| 16 | + | - ZFS ARC Size Floor: `zarcstat # c for floor` | |
| 17 | + | - ZFS ARC Size Ceiling: cat /sys/module/zfs/parameters/zfs_arc_max | |
| 18 | + | - ZFS ARC with alloc 50% of system mem by default, which could crash into a memory hugepage. | |
| 19 | + | - Balance Hugepages and ZFS ARC mem directly via the kernel command line | |
| 20 | + | ``` | |
| 21 | + | boot.kernelParams = [ | |
| 22 | + | # Balance Hugepage and ZFS ARC mem alloc limits directly via kernel command line | |
| 23 | + | "hugepages=96" | |
| 24 | + | "hugepagesz=1G" | |
| 25 | + | "default_hugepagesz=1G" | |
| 26 | + | "transparent_hugepage=never" | |
| 27 | + | "zfs.zfs_arc_min=4294967296" | |
| 28 | + | "zfs.zfs_arc_max=12884901888" | |
| 29 | + | ] | |
| 30 | + | ``` | |
Newer
Older